Research On Photovoltaic Energy Storage Configuration And Optimal Scheduling In The Whole County

With the promotion of the "dual carbon goal" strategy and the "county-wide photovoltaic" policy,distributed photovoltaics have entered a high-speed growth channel,supporting the "half of the sky" of new installations.However,distributed photovoltaics have problems such as scattered power station locations,large numbers,small single scale,and complex and diverse construction environments,which brings great challenges to promoting the friendly access of photovoltaic in the whole county.In order to provide high-quality electric energy to end users while ensuring the safe and reliable operation of the power system,it is necessary to accurately predict the photovoltaic power generation power in the whole county and reasonably configure the energy storage system to solve the shortcoming of randomness,intermittency and volatility of photovoltaic power generation,and provide strong support for the orderly promotion of the development of photovoltaic projects in the county.This paper focuses on three aspects: photovoltaic short-term power forecasting,county photovoltaic cluster power prediction and tracking the optimal configuration of energy storage for photovoltaic planned output,and the specific work content is as follows:(1)Aiming at the problem of low prediction accuracy of new county photovoltaic projects due to lack of data,a short-term prediction model of AM-GRU short-term photovoltaic power forecasting model based on data enhancement is proposed.Firstly,the WGAN-GP model is used to learn the distribution of real PV historical data samples,and high-quality new samples similar to the original samples are generated to enrich the dataset to achieve data enhancement.On the basis of PV data enhancement,the AM and GRU models are combined to enhance the processing ability of key information in the data samples while reducing the complexity of the model.The example results show that the proposed combined model can effectively improve the prediction accuracy of short-term photovoltaic power generation.(2)In view of the particularity of the contribution of the photovoltaic cluster in the whole county,a cluster superposition method considering spatial correlation is proposed to predict the cluster power of the photovoltaic in the whole county.Firstly,the condensation hierarchical clustering method based on EOF decomposition is used to divide the sub-regions of the whole county’s photovoltaics.Then,the Kendall rank correlation coefficient was used to analyze the similarity between the historical data of photovoltaic power generation power and the corresponding sub-regions of each power station,and several power stations with high correlation coefficients were selected as alternative benchmark power stations,and comprehensive comparison was made according to the prediction accuracy difference and administrative division of each power station,and finally the appropriate benchmark power station was selected.According to the predicted power and weight coefficient values of the benchmark power station,the PV predicted power of each sub-region is obtained by upscale,and finally the prediction results of all sub-regions are superimposed to obtain the PV cluster power prediction results of the whole county.The example results show that the proposed method can effectively improve the power prediction accuracy of photovoltaic clusters.(3)In order to ensure the friendly access of photovoltaic in the whole county,it is necessary to use energy storage to compensate for the prediction error of photovoltaic power generation,so as to track the output of photovoltaic power generation plan,so this paper proposes an optimal allocation method of photovoltaic energy storage in the whole county considering the uncertainty of photovoltaic power prediction error.Firstly,the prediction error of photovoltaic power generation power under different meteorological conditions is analyzed,and the uncertainty of the prediction error is described by introducing robust opportunity planning constraints.Then,with the goal of optimal energy storage configuration cost,an energy storage optimal configuration model is established to track the output of the county’s photovoltaic plan,and the convex approximation method is used to convert the modified model into a deterministic model for solving.The results show that the proposed method can maximize the economy of energy storage configuration while ensuring the compensation effect.